Schedule of Events | Symposia

Precision Networks Exhibit High Temporal Stability over Longitudinal Periods

Poster Session D - Monday, March 31, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom

Hyejin J. Lee¹, Ally Dworetsky², Alexis Porter³, Sihan Fei³, Benjamin A. Seitzman², Babatunde Adeyemo², Jessica R. Cohen⁴, Mark D'Esposito⁵, Maital Neta⁶, Steven E. Petersen², Caterina Gratton¹; ¹University of Illinois Urbana-Champaign, ²Washington University in Saint Louis, ³Northwestern University, ⁴University of North Carolina at Chapel Hill, ⁵University of California Berkeley, ⁶University of Nebraska-Lincoln

Measuring the system-level ‘network’ organization of the human brain can be useful for characterizing individual brain function and linking this function to phenotypic behavioral traits. A prerequisite for these applications is to understand how stable these measures are over time within individuals. Providing sufficient evidence that network features maintain high stability over significant periods could establish them as reliable tools for tracking individuals and detecting changes associated with psychiatric or neurodegenerative disorders. To address this need, we examined the temporal stability of brain networks in three participants over periods ranging from 6 to 13 years. We were able to reliably assess these measures for each individual using highly sampled precision resting-state fMRI data (>58 minutes per participant per time point, post-processed). The results showed stable network patterns, with consistent individual differences in the patterns even over a decade. We further conducted a systematic analysis of temporal stability by assessing functional connectivity, network assignments, and idiosyncratic network features in ten participants over 1-3 years. These measures demonstrated high consistency, underscoring the preservation of unique brain network features over time that define individual characteristics. Our findings support the use of brain network measures—provided sufficient participant data are collected—for phenotypic prediction and biomarker research.

Topic Area: METHODS: Neuroimaging